Christopher Pedler

STUDIES ON DEVELOPING RETINAL VESSELS: IX. REACTION OF ENDOTHELIAL CELLS TO OXYGEN*

IN reviewing the mechanisms which might be involved in the phenomenon of oxygen vaso-obliteration in the immature retina, it was recently suggested that swelling of the retinal tissue around the vessels could be the primary factor concerned. The detailed evidence upon which this suggestion was based has already been reported (Ashton, 1957; Ashton, Graymore, and Pedler, 1957) and further supporting data have since been published (Graymore, 1958; Pedler, 1959a, b). It was pointed out, however, that a more direct action of oxygen on the vessel wall could not be excluded by the negative evidence then obtained, because the techniques available at that time were not entirely satisfactory. With the development of the new digest method of Kuwabara and Cogan (1960), it has been possible to isolate the affected capillaries, and with the icquisition of an electron microscope at this Institute a detailed study of the changes in the internal structure of the endothelial cell has become feasible. The findings obtained are presented and discussed in this paper.

The fine structure of the radial fibres in the reptile retina

The radial fibre system in the reptilian retina is particularly profuse and shows a number of unusual morphological features which may throw some light on its function. Retinas from Hemidactylus turcicus, Phelsuma inunguis, Hemidactylus flaviviridius, Gecko gecko , and Lacerta muralis were examined, using osmium-fixed and aralditeembedded tissue. It is suggested that the complex specialization of structure in the vicinity of the visual cell surface has developed to facilitate interchange of fluids or metabolites across the visual cell membrane in the absence of any intraretinal vessels. It is also suggested that the complex interlocking of the radial fibre processes, particularly in the nerve fibre layer, is probably a means of preserving precise inter-relationships of neuronal material to minimize relative movement in the interests of visual acuity. It is speculated that the profuse glial specialization found on the vitread surface of the cone pedicle may represent one of the pathways for lateral inhibitory impulses passing between adjacent visual cells, particularly since no evidence of synaptic components was found in the cytoplasm of the so-called horizontal cells which are usually accepted to be the structural pathway for the lateral passage of information in the retina. A morphological discrepancy between the light and the electron microscopic appearances of the radial fibre system is discussed, and an explanation offered. Finally, a new retinal layer is described.

The fine structure of the cone of a diurnal gecko (Phelsuma inunguis)

Using material fixed in osmium tetroxide, block stained with phosphotungstic acid and embedded in Araldite, the fine structure of the cone of a diurnal gecko (Phelsuma inunguis) has been studied. The three basic cell varieties proposed by Underwood have been confirmed, and the two classes of double cell have been shown to be pairs of intact single cells with parts of their plasma membranes in very close apposition. The outer segment, inner segment, myoid region and nuclear zone are the parts involved in the double relationship, the synaptic pedicles in both classes being entirely separate. At the junction of the inner and outer segments, no clearly defined cilial structures were encountered and abnormal lamellae were often seen. The contents of the inner segment are described and the relationship of the oil droplet present in the smaller member of the Class B double to its surrounding mitochondria is shown. The fine structure of the paraboloid is also described and its probable derivation from the endoplasmic reticulum suggested. The synaptic pedicle is unusually large and separated from the nuclear zone by a narrow constriction, no true conducting fibre being evident. Components from four cells were found to be in close relationship to the proximal end of the pedicle: the horizontal cells, bipolar cells, radial fibres of Muller and probably processes from adjacent cones.

UNUSUAL COLOBOMA OF THE OPTIC NERVE ENTRANCE

DEVELOPMENTAL anomalies of the eye in the region of the optic disc are commonly divided into two groups: those in which failure of coaptation occurs within the optic nerve sheath and those showing lesions of the peripapillary choroid or sclera. With the exception of the small holes sometimes seen on the anterior surface of the optic disc, which are probably related to aberrant formation of Bergmeisters papilla, the first type is much the less frequent of the two, and probably all the anomalies in the second group are formed by imperfect closure of the foetal fissure with the result that they appear on a line running inferiorly from the rim of the optic disc to the equator. The following report describes a coloboma of the optic nerve entrance of unusual configuration which does not belong to either group.

The fine structure of the tapetum cellulosum.

This paper reports detailed observations on the fine organization of the tapetal cell in the cat, describes some changes in structure produced by immersion in solutions of varying tonicity and discusses their relevance to the mechanism of tapetal reflection.

The fine structure of the corneal epithelium.

The fine structure of the corneal epithelium has been investigated with particular reference to the intercellular attachment areas or desmosomes. It has been found that the desmosomes vary markedly in structure; those on the basement membrane being different from those between the wing cells, and those between the wing cells again showing differences when compared with the connections between the superficial cells. The cytoplasmic organelles of the basal cells are shown to be aggregated on the immediately posterior aspect of the nuclear membrane, an observation which may usefully be correlated with the known fact that ocular virus inclusion bodies are commonly seen as a perinuclear cap of similar outline, suggesting that certain viruses may develop more readily in the presence of such organelles. The anterior surface of the cornea is shown to possess blunt cytoplasmic protrusions, which may account for the ability of the cornea to remain moist for long periods.

STUDIES ON DEVELOPING RETINAL VESSELS V. MECHANISM OF VASO-OBLITERATION (A Preliminary Report)

SINCE the discovery of the specific effect of oxygen on the immature retina, which manifests itself in vasoconstriction leading to total obliteration of the ingrowing vascular complexes (Ashton, Ward, and Serpell, 1953, 1954), the work of this department has been concerned with experiments designed to elucidate the mechanism of this effect. The evidence that the phenomenon is entirely confined to vessels actually growing in the retina, and then only when the retina is immature and in its normal anatomical position, has been recently reviewed elsewhere (Ashton, 1957). Broadly speaking there would appear to be three main theoretical explanations for the closure of these vessels:

STUDIES ON DEVELOPING RETINAL VESSELS VII. FLUORIDE-INDUCED VASO-OBLITERATION AND ITS RELATION TO RETINAL MATURITY

SINCE it was first reported that oxygen had a specific vaso-obliterative action on the vessels of the immature retina (Ashton, Ward, and Serpell, 1953), a working hypothesis has been devised and certain experiments carried out in an attempt to understand the intermediate mechanism involved (Ashton, Graymore, and Pedler, 1957). In the latter paper it was suggested that the vessels might close through external pressure from fluid imbibition into the retina as a result of an interruption in glycolysis brought about by ambient hyperoxia. Experiments were also reported wherein vessel closure analogous to that following oxygen exposure could be induced by the intravitreal administration of established inhibitors of glycolysis, such as sodium fluoride, which was found to produce complete retinal vaso-obliteration in a consistently repeatable manner. This finding alone is not sufficient to support the idea of cell swelling as a cause of vaso-obliteration in the immature retina, and many other facets of this hypothesis still require experimental investigation; however,, support has recently been provided by experiments (Graymore, 1958) which showed that, under defined conditions in vitro, excised retinae incubated in solutions containing known amounts of sodium fluoride underwent a significant increase in Wet Weight: Dry Weight ratio. Similarly, other experiments have demonstrated that the intravitreal injection of sodium fluoride into the intact living eye can result in a histologically measurable swelling of the inner cell layers of the immature but not of the mature retina (Pedler, 1959). Hence, if retinal swelling is responsible for the vaso-obliteration observed in the living eye, then the vessel closure should also show correlation with the degree of retinal maturity. This paper reports the results of some experiments designed first to investigate this possibility, and secondly to determine whether there is any similarity between the age at which fluoride ceases to affect the retinal vessels and that at which ambient hyperoxia becomes ineffective.

STUDIES ON DEVELOPING RETINAL VESSELS VI. HISTOLOGICAL MEASUREMENT OF FLUORIDE-INDUCED SWELLING IN THE RETINA

IT is now a firmly established experimental finding that the vessels of the -immature retina specifically close as a result of raising the ambient oxygen concentration. Since this was first reported (Ashton, Ward, and Serpell, 1953) subsequent research has been directed towards the elucidation of the underlying mechanism. Some progress has been made, and it has recently been reported (Ashton, Graymore, and Pedler, 1957) that when certain established enzyme inhibitors were injected through the vitreous of an immature animal on to the retinal surface, complete obliteration of the retinal vasculature occurred within a short space of time. This effect was studied in the living animal, using a direct observation technique (Ashton and Cook, 1954; Pedler, 1957), and on repetition was found to be consistent, not only in the time elapsing between injection and vaso-obliteration, but also in the order of closure of the vessels concerned. These experiments have led to the formulation of a working hypothesis that the vessel closure observed might be caused by external pressure resulting from swelling of surrounding retinal tissue. This possibility has been discussed (Ashton and others, 1957) and further work (Graymore, 1958) has provided some support for it by showing that when the isolated retina is incubated in solutions containing sodium fluoride a significant increase in Wet Weight: Dry Weight ratio takes place. It was thought possible that such an intake of fluid might result in an actual increase in thickness which would be measurable histologically. This paper reports the results of some experiments performed to investigate this possibility.

STUDIES ON DEVELOPING RETINAL VESSELS: IV. EFFECT OF IONIZING RADIATION

THE EFFECT of hyperoxia on the vasculature of the immature retina is now well established, and it has been repeatedly demonstrated that when a newlyborn kitten is subjected to an ambient oxygen concentration exceeding approximately 35 per cent. a total and permanent vaso-obliteration will develop within a period of 8 to 36 hours (Ashton, Ward, and Serpell, 1954). This observation has provided the basis for much of the significant work carried out on the aetiology of retrolental fibroplasia, and may in fact provide useful information about other diseases of the retina. In view of this, it is clearly essential to assemble more information about the mechanism by which oxygen is able to produce vaso-obliteration in the immature retina. In recent years a considerable amount of work has been published suggesting that the toxic effects of both oxygen and ionizing irradiation may be brought about by the inhibition or destruction of common enzymatic systems. For example, reduction of metabolism in the experimental animal by cooling, provides some protection again oxygen poisoning, whereas raising the body temperature enhances oxygen susceptibility (Grossman and Penrod, 1949; Stadie, Riggs, and Havgaard, 1944). Similarly it appears that a lowered metabolism decreases the amount of damage incurred by a given dose of radiation (Ord and Stocken, 1953). Furthermore, both agents affect some of the enzymes of intermediary metabolism in a comparable fashion; for instance, the sulphydryl enzymes, pyruvic oxidase and succinate dehydrogenase, are both inactivated by high oxygen concentration (Dickens, 1946), and the activity of diphosphoglyceric-aldehyde dehydrogenase is almost completely destroyed in vitro by a radiation exposure of 500 r. (Barron, Dickman, Muntz, and Singer, 1949). These reactions are very probably produced by the interaction of various oxidizing free-radicals transiently elaborated by ionising radiation from the breakdown of water, for there is considerable evidence to show that irradiation acts indirectly by forming toxic peroxide and hydroperoxyl ions (Rajewsky, 1952; Dainton, 1951). In fact the yield of these substances is very considerably enhanced by increasing the quantities of molecular oxygen present; indeed, the incidence of chromosomal aberrations in inflorescences of algae can be made to rise as the oxygen concentration increases (Giles and Riley, 1950), and certain animal tumours can be made as much as three times more radiosensitive if they are irradiated in atmospheres of high oxygen concentration (Gray,